ARTICLES
PUBLISHED ONLINE: 24 FEBRUARY 2013 | DOI: 10.1038/NGEO1738
Antarctic Bottom Water production by intense
sea-ice formation in the Cape Darnley polynya
Kay I. Ohshima1 *† , Yasushi Fukamachi1† , Guy D. Williams2† , Sohey Nihashi3 , Fabien Roquet4 ,
Yujiro Kitade5 , Takeshi Tamura6 , Daisuke Hirano5 , Laura Herraiz-Borreguero2 , Iain Field7 ,
Mark Hindell8 , Shigeru Aoki1 and Masaaki Wakatsuchi1
The formation of Antarctic Bottom Water—the cold, dense water that occupies the abyssal layer of the global ocean—is
a key process in global ocean circulation. This water mass is formed as dense shelf water sinks to depth. Three regions
around Antarctica where this process takes place have been previously documented. The presence of another source has
been identified in hydrographic and tracer data, although the site of formation is not well constrained. Here we document
the formation of dense shelf water in the Cape Darnley polynya (65◦ –69◦ E) and its subsequent transformation into bottom
water using data from moorings and instrumented elephant seals (Mirounga leonina). Unlike the previously identified sources
of Antarctic Bottom Water, which require the presence of an ice shelf or a large storage volume, bottom water production
at the Cape Darnley polynya is driven primarily by the flux of salt released by sea-ice formation. We estimate that about
0.3–0.7 × 106 m3 s−1 of dense shelf water produced by the Cape Darnley polynya is transformed into Antarctic Bottom Water.
The transformation of this water mass, which we term Cape Darnley Bottom Water, accounts for 6–13% of the circumpolar total.
A
ntarctic bottom water (AABW) is the cold, dense water
in the abyssal layer, accounting for 30–40% of the global
ocean mass1 . AABW production is a major contributor to
the global overturning circulation1–3 and represents an important
sink for heat and possibly CO2 (ref. 4). AABW originates as
dense shelf water (DSW), which forms on the continental shelf
by regionally varying combinations of brine rejection from sea-ice
growth and ocean/ice-shelf interactions. With sufficient negative
buoyancy and an export pathway across the shelf break, the
DSW can mix down the continental slope with ambient water
masses to produce AABW (ref. 5). At present, AABW production
is recognized in three main regions: the Weddell Sea6–9 , the
Ross Sea10–12 and off the Adélie Coast13–15 (Fig. 1, inset map).
In the Weddell and Ross seas, large continental embayments
associated with major continental ice shelves were considered
necessary mechanisms to generate sufficient negative buoyancy
in local DSW for the production of AABW (refs 9,10). This
paradigm was broken when Adélie Land Bottom Water was
directly linked to coastal polynyas, thin ice areas of enhanced
sea-ice production and associated brine rejection. Although
many polynya regions exist, particularly in East Antarctica, the
storage capacity of the continental shelf in this region was
proposed as the reason this seemed to be the only polynya
region capable of forming sufficiently saline DSW for producing
AABW (refs 14,15).
Another independent variety of AABW, previously identified
in the Weddell–Enderby Basin (Fig. 1, inset map) from offshore
properties detailed in hydrographic and tracer studies16–21 , has
presented a puzzle in terms of its existence and DSW source. The
Prydz Bay region (71◦ –80◦ E) seemed the most likely candidate22–24 ,
with its large continental embayment and the presence of the
Amery Ice Shelf resembling the Weddell Sea and Ross Sea AABW
regions. However, the results from several ship-based studies21 were
inconclusive and the export of DSW and its downslope transport
were never observed from this region.
Most recently, satellite-derived estimates of sea-ice production
suggest that the Cape Darnley polynya (CDP), located northwest
of the Amery Ice Shelf (Fig. 1), has the second highest ice
production around Antarctica after the Ross Sea Polynya25 . This
satellite analysis provided the motivation for an extensive Japanese
mooring programme conducted in 2008–2009 as part of the
International Polar Year, to prove the production of AABW
from the CDP. Here we present the successful results of the
moorings and observations of new AABW on the continental
slope off the CDP. Furthermore, we confirm that the CDP is the
DSW source for this AABW based on instrumented seal data,
which has recently become an important source of hydrographic
profiles in logistically challenging regions/seasons around the
Antarctic margin26–28 . Finally, the contribution of this AABW to
total AABW is discussed.
Intense sea-ice production in the polynya
A new high-resolution satellite data set is used here to enhance
understanding of the CDP from previous studies25,29,30 . A typical
synthetic aperture radar image of the CDP (Fig. 1) clearly reveals
the extent of the CDP (>104 km2 ), in which the newly formed sea
ice appears as white streaks (high radar backscatter). The annual
ice-production estimate from the Advanced Microwave Scanning
Radiometer-EOS (AMSR-E) data (Supplementary Section S1)
details the extremely high ice production of 195 ± 71 km3 , with
1 Institute
of Low Temperature Science, Hokkaido University, Sapporo 060-0819, Japan, 2 Antarctic Climate and Ecosystem Cooperative Research Centre,
University of Tasmania, Hobart 7001, Australia, 3 Tomakomai National College of Technology, Tomakomai 059-1275, Japan, 4 Department of Meteorology,
Stockholm University, S-106 91 Stockholm, Sweden, 5 Tokyo University of Marine Science and Technology, Tokyo 108-8477, Japan, 6 National Institute of
Polar Research, Tachikawa 190-8518, Japan, 7 Graduate School of the Environment, Macquarie University, 2109, Australia, 8 Institute for Marine and
Antarctic Studies, University of Tasmania, 7001, Australia. † These authors contributed equally to this work. *e-mail: [email protected].
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
235
NATURE GEOSCIENCE DOI: 10.1038/NGEO1738
ARTICLES
W
(dB)
¬10
¬5
Ca
ny
on
0
¬15
0
3,0
¬20
ild
00
1,0 00
5
M4
B1
0.1 m s¬1
50 km
Daly
3,000
2,500
Cany
on
Prydz
Bay
M3
66° S
M1
2,5
0
0
67° S
B5
M2
Weddell¬Enderby Basin
Weddell Sea
B.
Amery Ice Shelf
.
B
N.
Antarctica
Ross Sea
68° S
B.
Cape Darnley
t
oas
lie C
Adé
135° W
Antarctica
68 oE
66 oE
66° E
135° E 64° E
68° E
70° E
Figure 1 | Intense sea-ice production in the CDP, revealed from satellite data. The background ENVISAT ASAR image is from 7 August 2008. The
black–white graded scale bar in the top left indicates the backscatter coefficient. Annual sea-ice production in 2008, estimated from the AMSR-E data, is
indicated by red (>10 m yr−1 ) and pink (>5 m yr−1 ) contours. Fast ice areas, as detected by the AMSR-E algorithm, are highlighted in green. The
bathymetry is indicated by blue contours. Mooring locations are indicated by orange symbols, with the mean velocity at the bottom and top of the mooring
shown as orange and yellow vectors, respectively. CTD station lines A, B and C are denoted by squares, circles and crosses, respectively. The Nielsen (N.B.)
and Burton (B.B.) basins are as shown. The inset map shows the location of the study area (red rectangle).
Ice prod.
(km3)
a
1.5
0.0
08/F
b
M
A
M
J
J
A
S
O
N
D
09/J
J
J
A
S
O
N
D
09/J
θ (°C)
0.0
¬1.5
M3 (bottom depth: 2,608 m)
08/F
c
M
A
M
28.5
γn
Salinity
34.7
34.6
M
A
M
J
J
A
S
O
N
D
09/J
08/F
M
A
M
J
J
A
S
O
N
D
09/J
0.5
0.0
500
0.15
(3)
0
0.00
08/F
M
A
M
J
J
A
Month (2008¬09)
S
O
N
D
Layer (m)
e
Transport
(Sv km¬1)
Velocity
(m s¬1)
d
28.3
08/F
09/J
Figure 2 | Observations of new AABW production in Wild Canyon offshore from the CDP. a, Daily sea-ice production in the CDP estimated from AMSR-E
data and a heat-flux calculation. b–e, M3 mooring time-series of potential temperature (θ ) at depths of 26 m (blue) and 224 m (red) from the bottom (b),
salinity (orange) and neutral density (γ n , green) at 26 m from the bottom (c), the velocity component of the mean (dominant) flow direction at 20 m
(blue) and 226 m (red) from the bottom (d) and the estimated layer thickness (red, 15-day running mean) and volume transport per unit width (light blue,
daily mean; dark blue, 15-day running mean) of new AABW (e). The value in parentheses on the left axis shows the volume transport (Sv = 106 m3 s−1 ) for
an assumed current width of 20 km in e.
236
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
M2 2,000
M3
2,500
M4
2,000
00
1,5,0
0
1 0
67° S
51,00000
50
3,00
0
0
500
200
500
200
0
2,5
ARTICLES
2,0
1,5000
510,000
0 0
M1
2,50
0
66° S
0
3,00
a
2,500
NATURE GEOSCIENCE DOI: 10.1038/NGEO1738
0
200
500
68° S
500
50
0
69° S
200
50
0
66° E
69° E
34.4
b
500
72° E
34.5
34.6
75° E
78° E
34.7
34.8
Salinity
35.0
34.8
34.6
34.4
M
A
2,000
2,500
66.5° S
M
M3
J
J
A
00
F
M4
2,0
11/J
c
00
1,5
0
,0 0
00
1
0
1,00
50
67.0° S
34.60
69° E
34.52
0
20
200
200
68° E
34.64
34.56
B5
67° E
34.68
0
2,0
0
1,50
500
S
70° E
71° E
Figure 3 | Bottom salinities of DSW and mSW from the instrumented
seals. a, Salinities of bottom-of-dive data for all available seal profiles with
the bottom depth >250 m and potential temperature θ < −1.7 ◦ C over the
continental shelf off Cape Darnley and in Prydz Bay. All other dive
locations are indicated by small grey points. It is noted that, for these
temperatures and depths, salinity >34.5 approximately corresponds to
γ n > 28.27 kg m−3 , threshold of AABW density2 . b, Spatial and temporal
variation in salinities of bottom-of-dive data shown in a. Symbols follow
those in a, coloured according to the regional outlines in a. c, Bottom
salinities for overflows of mSW (θ < −0.8 ◦ C, bottom depth >800 m) in
the region between 67◦ and 71◦ E. The annual sea-ice production is
contoured as in Fig. 1.
local production rates exceeding 10 m yr−1 . The dark zone with
small white patches just east of the polynya is a grounded iceberg
tongue30 (highlighted in green). With the prevailing ocean currents
directed westward and dominant offshore winds, newly formed
sea ice accumulates to the east of this barrier and is advected
away to the west. This mechanism makes this polynya the second
largest in area in the Antarctic, with a large associated salt flux
(Supplementary Table S1).
New AABW production off the polynya
The moorings were deployed at four locations: at M1 (water depth:
2,923 m) and M3 (depth: 2,608 m) within the Daly and Wild
canyons, predicted to be the DSW/AABW pathways, and at M2
(depth: 1,439 m) and M4 (depth: 1,824 m) over the continental
slope (see Fig. 1 and Supplementary Table S2). M3, located at
the centre of Wild Canyon just off the CDP, showed the most
prominent downslope current (orange vector in Fig. 1). The
mean flow was directed downslope, as determined by multibeam bathymetry measurements during deployment, and had
prominent bottom intensification (Fig. 1). In May, two months
after the onset of active sea-ice production (Fig. 2a), a colder,
less saline, and denser signal abruptly appeared at all instruments
(up to 224 m from the bottom) and became dominant after June
(Fig. 2b,c). The current speeds across this layer intensified in
conjunction with the cold and dense signal (Fig. 2d). Thereafter,
downslope current speeds of up to 0.5 m s−1 fluctuated coherently
with these water properties. Spectral analysis reveals strong peaks
at periods of 4 and 5 days. The thickness of the cascading
AABW layer was estimated by linear extrapolation to increase
up to ∼400 m from May to October and to remain >300 m
to January (Fig. 2e and Supplementary Section S5a). With these
properties (neutral density γ n > 28.27 kg m−3 ) at this depth
range (>2,500 m), these observations represent new AABW that
would be within the range of AABW in the Weddell Sea2
(Supplementary Fig. S8a).
We consider the observed downslope flows, with cold, dense
properties and strong bottom intensification, to be gravity
currents (Supplementary Section S2c). Together with analysis of
a salinity budget estimated from the satellite data (Supplementary
Section S6a), we propose that after two months of high sea-ice
production in the CDP, the salinity and hence density of the shelf
water is sufficient to descend the slope, channelled by the canyons.
The fluctuating signal at M3 probably originates from a periodic
outflow due to the baroclinic instability associated with the front
between this DSW and the less dense ambient water offshore31 .
Thereafter, the DSW can descend in the form of an eddy or
plume5,31–33 . The cold dense water could descend to even greater
depths with the aid of thermobaricity9,34 .
The moorings at M1, M2 and M4 also reported varying degrees
of new AABW production. Cold and dense water was observed
at moorings M1 and M2 shortly after the onset of active sea-ice
production (Supplementary Fig. S3a–d). These data indicate that
DSW from the CDP is initially just advected westwards along the
slope (M2) and that new AABW is also transported down Daly
Canyon (M1). A cold and dense signal was also found at M4
(Supplementary Fig. S3e,f) and possibly indicates the influence of
DSW formed upstream of the CDP. However, the near-bottom
signal is much thinner (Supplementary Fig. S5), suggesting that its
contribution is minor relative to the CDP.
DSW revealed by the instrumented seals
Southern elephant seals (Mirounga leonina) equipped with
conductivity–temperature–depth (CTD) sensors (Supplementary Section S4) reveal the spatial distribution and seasonal
evolution of DSW properties over the continental shelf off
Cape Darnley and to the east in Prydz Bay. Hydrographic data
were post-processed using a set of delayed-mode techniques35 ,
yielding sufficient accuracies for the characterization of DSW,
that is, about 0.03 ◦ C in temperature and 0.05 or better in
salinity. The bottom-of-dive salinity (Supplementary Section S4c)
for all available seal profiles with bottom depth >250 m and
potential temperature θ < −1.7 ◦ C is shown in Fig. 3a,b. The
spatial distribution clearly reveals the high-salinity DSW over the
CDP region (Fig. 3a), consistent with the high ice production
(Fig. 1). We identify six regional subsets that cover the continental shelf in this region (Fig. 3a) and find the most saline
DSW in the inshore region of the continental shelf west of
Cape Darnley (red outline in Fig. 3a), with actively forming
(late June through early July) and remnant (January–February)
DSW salinities >34.8 (red data in Fig. 3b). These are among
the most saline and therefore densest shelf waters observed
around Antarctica14,15,24,36,37 . This region overlaps with the high
annual ice production of >5 m yr−1 (Fig. 1). In contrast, there
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
237
NATURE GEOSCIENCE DOI: 10.1038/NGEO1738
ARTICLES
N
Wild Canyon
0
¬1,000
¬3,000
¬4,000
Depth (m)
¬2,000
Daly Canyon
New AABW
Figure 4 | Schematic of CDBW production. Enhanced sea-ice production
(red shading) in the CDP formed leeside of the grounded iceberg tongue
causes DSW formation through the high brine (salt) rejection. The
downslope pathway from DSW to new AABW is represented by the purple
arrows. The primary DSW descends down Wild Canyon. The remainder
flows westwards over the shelf and slope with geostrophic balance and
then downslope elsewhere in the west (mostly along Daly Canyon). DSW,
through mixing with overlying CDW, is gradually transformed into new
AABW (CDBW). B.B., Nielsen basin; N.B, Burton basin.
is a lower-salinity (34.5–34.6) variety of DSW observed over
the Prydz Bay shelf from March to mid-June (green, blue and
purple data in Fig. 3b).
Several of the instrumented seals foraged on the continental
slope to depths as great as 1,800 m. These returned very rare
wintertime measurements of overflowing DSW, also referred to
as modified Shelf Water15,37 (mSW), in the region just off CDP
and southeast of M3, between 67.5◦ and 71.0◦ E. The bottom
salinities from available profiles with strong mSW signatures
(colder, fresher and denser relative to the ambient slope water,
see Supplementary Fig. S6) are shown in Fig. 3c. The key feature
here is the clear shift to more saline and denser (>34.64)
bottom values from east to west, in particular west of ∼68.5◦ E.
These results, along with the mooring results, confirm that
high-salinity DSW from the CDP region is the main source of
the new AABW cascading down the canyons. Although there
is likely to be some weak contribution or pre-conditioning
by low-salinity DSW exported from Prydz Bay, it cannot be
quantified at this point.
Implications of Cape Darnley Bottom Water production
Our discovery of new AABW production offshore from the CDP
is a major missing piece of a 30-year-old puzzle regarding the
source of recently ventilated AABW in the Weddell–Enderby Basin.
We propose to name this AABW Cape Darnley Bottom Water
(CDBW) and summarize its production schematically in Fig. 4.
DSW produced in the CDP descends down Wild and Daly canyons
and is transformed into CDBW through mixing with overlying
Circumpolar Deep Water (CDW). On the basis of its observed
properties, relative to the water masses reported in the Weddell
Sea18–20 , CDBW should migrate further westward, and increase
its volume by gradual mixing with CDW, to ultimately constitute
part of the AABW in the Weddell Sea referred to as Weddell Sea
Deep Water (WSDW).
We now investigate the DSW ventilation rate, which is the
volume flux of DSW exported from the continental shelf that
ultimately produces CDBW, and thereafter the contribution
of CDBW to total AABW. The annual transport of CDBW
238
down Wild Canyon is estimated from mooring data at M3 to
be 0.52 ± 0.26 Sv (Sv = 106 m3 s−1 ; Methods). In conjunction
with another estimate using a salinity budget from satellite
data, we propose that 0.3–0.7 Sv of DSW is ventilated in this
region (Methods). This represents ∼6–13% of the circumpolar
contribution based on the total DSW ventilation rate of 5.4 Sv
determined from chlorofluorocarbon data38 and is of the same
order of magnitude as DSW export reported from Adélie Land14 .
This significant surface-to-bottom injection is consistent with the
local maxima of near-bottom chlorofluorocarbon and oxygen
concentrations offshore and to the west of Cape Darnley, as
shown in the detailed hydrographic atlas20 . To examine how
much WSDW is renewed through CDBW, we adopt the mean
property of AABW in the Atlantic sector2 , which is close to
that of WSDW (Supplementary Fig. S8a). Using the DSW
ventilation rate estimated above and the mixing ratio of CDBW
and CDW to this mean property, the contribution of CDBW
is estimated to be 0.65–1.5 Sv (Methods). There have been
several estimates2,39,40 for total AABW production in the Atlantic
sector, in the ∼3–10 Sv range. Taking the well-referenced value
of 4.9 Sv (ref. 2), the CDBW contribution is ∼13–30% of the
Atlantic AABW production.
In comparison to the other AABW formation regions, we note
that the periodicity of the overflows is similar to that reported
in the Weddell32 and Ross34 seas and the AABW layer thickness
is comparable to that in the Ross Sea11,12 . One difference is
that the influence of tides is much weaker here (<0.05 m s−1 )
relative to the Ross Sea11,12 . Most importantly, the Cape Darnley
region demonstrates conclusively that a relatively narrow section
of continental shelf with limited DSW storage capacity can produce
AABW from polynya-driven sea-ice production alone. We therefore
speculate that there could be further AABW-formation discoveries
in similar polynyas, particularly those in East Antarctica15,25
(Supplementary Table S1).
High sea-ice production in the CDP results from Cape Darnley
ice barrier (iceberg tongue) blocking the westward advection of
sea ice. This mechanism is similar to that of the Mertz Glacier
Tongue and Drygalski Ice Tongue on the Mertz and Terra Nova
Bay polynyas, respectively. As demonstrated by the calving of
Mertz Glacier Tongue in 2010, the rapid change (collapse) of
such a barrier can cause a pronounced change in sea-ice and
AABW production41,42 . Although the iceberg tongue and sea-ice
production in the CDP were relatively stable during the AMSRE period from 2003 to 2010 (Supplementary Section S1), any
significant change to the structure of the ice barrier would strongly
impact CDBW production.
This study now raises the profile of CDBW to that of an
important contributor to WSDW, a major component of the
AABW driving the lower limb of the meridional overturning
circulation (MOC) of the Atlantic Sector43 . It has been reported
that WSDW has been warming since the 1980s (ref. 44), with
its volume possibly contracting45 , and that this could result in
a weakening of the MOC. Furthermore, sediment-core records
taken around the CDP indirectly suggest that there has been
millennium-scale variability in the local AABW production46 . It
is vital that CDBW be incorporated into the global assessment
of the MOC, a key element of the climate system. This will
improve numerical simulations predicting its response to longterm climate change.
Methods
Estimation of the DSW ventilation rate. We make two estimates of the DSW
ventilation rate. The first estimate is based on the transport of AABW down
Wild Canyon using velocity and temperature data from M3 (Supplementary
Section S5). New AABW is defined as water with potential temperature
θ < −0.4 ◦ C, based on the bimodal distribution of near-bottom temperature
that separates around −0.4 ◦ C (Supplementary Fig. S7). The AABW layer
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
NATURE GEOSCIENCE DOI: 10.1038/NGEO1738
thickness and speed are inter/extra-polated from the available instrument
range. The layer thickness and velocities from this single mooring provide the
transport of new AABW (CDBW) per unit width (Fig. 2e). When the width
of the downslope current is assumed to be 20 km based on the canyon width,
the annual transport of CDBW is 0.52 ± 0.26 Sv. Using this transport, the
DSW ventilation rate is estimated to be 0.30 ± 0.15 Sv on the assumption that
CDBW (θ = −0.65 ◦ C, S = 34.635) is a 42:58 mixture of CDW (θ = 1.05 ◦ C,
S = 34.71) to DSW (θ = −1.9 ◦ C, S = 34.58) from observed water properties
(Supplementary Fig. S8a).
The second estimate of the DSW ventilation rate is based on a salt budget
from sea-ice production (Supplementary Section S6b), where we assume that
summer water masses are directly transformed into winter DSW by the salt flux
from sea-ice production in the CDP. Using the annual sea-ice production derived
from the satellite data and heat-flux calculation (Supplementary Fig. S2), the
salt-budget calculation yields an annual DSW ventilation rate of 0.70 ± 0.25 Sv.
The 0.30 Sv estimated from the CDBW transport down Wild Canyon represents
a lower bound for the entire CDP region because there is further CDBW
transport mainly down Daly Canyon. Meanwhile, the 0.70 Sv derived from
the salt budget represents an upper bound, because the sea-ice production
in the CDP does not directly equate to DSW formation and export for
CDBW production.
Estimation of the CDBW contribution to AABW production. The AABW
produced off the CDP (CDBW) is further mixed with overlying warmer water to
become general AABW in the Weddell Sea. We evaluate the flux of CDBW that
contributes to the final output of AABW in this sector. When the averaged property
of AABW in the Atlantic sector2 (mainly in the Weddell Sea; θ = −0.30 ◦ C,
S = 34.66) is plotted in the θ–S diagram of Supplementary Fig. S8a, it is located near
the mixing line linking DSW, CDBW and CDW. Assuming that CDBW is finally
mixed with the CDW to become the averaged AABW, the mixing ratio of CDBW
and CDW is calculated to be 79:21, based on θ. Using this mixing ratio and the
estimated annual transport of CDBW down Wild Canyon (0.52 Sv), the ultimate
contribution of CDBW to the AABW in the Weddell Sea would be 0.65 Sv. Further,
if we use this mixing line and ratio for the original end point of DSW and apply
them to the upper bound estimation of DSW from the sea-ice production, the
ultimate contribution of CDBW would be 1.5 Sv (=0.65×0.7/0.3 Sv).
Animal ethics. Animal ethics for all animal handling was approved by
the University of Tasmania and Macquarie University’s Animal Ethics
Committees.
Received 25 June 2012; accepted 21 January 2013; published online
24 February 2013
References
1. Johnson, G. C. Quantifying Antarctic Bottom Water and North Atlantic Deep
Water volumes. J. Geophys. Res. 113, C05027 (2008).
2. Orsi, A. H., Johnson, G. C. & Bullister, J. L. Circulation, mixing, and
production of Antarctic Bottom Water. Prog. Oceanogr. 43, 55–109 (1999).
3. Marshall, J. & Speer, K. Closure of the meridional overturning
circulation through Southern Ocean upwelling. Nature Geosci. 5,
171–180 (2012).
4. Sigman, D. M. & Boyle, E. A. Glacial/interglacial variations in atmospheric
carbon dioxide. Nature 407, 859–869 (2000).
5. Baines, P. G. & Condie, S. in Ocean, Ice, and Atmosphere: Interactions at the
Antarctic Continental Margin (eds Jacobs, S. & Weiss, R.) 29–49 (Antarct. Res.
Ser., Vol. 75, American Geophysical Union, 1998).
6. Foster, T. D. & Carmack, E. C. Frontal zone mixing and Antarctic Bottom
Water formation in the southern Weddell Sea. Deep Sea Res. Oceanogr. Abstr.
23, 301–317 (1976).
7. Gordon, A. L., Huber, B. A., Hellmer, H. H. & Ffield, A. Deep and bottom
water of the Weddell Sea’s western rim. Science 262, 95–97 (1993).
8. Fahrbach, E., Harms, S., Rohardt, G., Schröder, M. & Woodgate, R. A.
Flow of bottom water in the northwestern Weddell Sea. J. Geophys. Res. 106,
2761–2778 (2001).
9. Foldvik, A. et al. Ice shelf water overflow and bottom water formation in the
southern Weddell Sea. J. Geophys. Res. 109, C02015 (2004).
10. Jacobs, S. S., Amos, A. F. & Bruchhausen, P. M. Ross Sea oceanography
and Antarctic Bottom Water formation. Deep-Sea Res. Oceanogr. Abstr. 17,
935–970 (1970).
11. Whitworth, T. III & Orsi, A. H. Antarctic bottom water production and export
by tides in the Ross Sea. Geophys. Res. Lett. 33, L12609 (2006).
12. Gordon, A. L. et al. Western Ross Sea continental slope gravity currents.
Deep-Sea Res. II 56, 796–817 (2009).
13. Rintoul, S. R. in Ocean, Ice, and Atmosphere: Interactions at the Antarctic
Continental Margin (eds Jacobs, S. & Weiss, R.) 151–171 (Antarct. Res. Ser.,
Vol. 75, American Geophysical Union, 1998).
ARTICLES
14. Williams, G. D., Bindoff, N. L., Marsland, S. J. & Rintoul, S. R. Formation
and export of dense shelf water from the Adélie Depression, East Antarctica.
J. Geophys. Res. 113, C04039 (2008).
15. Williams, G. D. et al. Antarctic Bottom Water from the Adélie and George V
Land coast, East Antarctica (140–149◦ E). J. Geophys. Res. 115, C04027 (2010).
16. Jacobs, S. S. & Georgi, D. T. in A voyage of discovery (ed. Angel, M.) 43–84
(Pergamon Press, 1977).
17. Mantisi, F., Beauverger, T., Poisson, A. & Metzl, N. Chlorofluromethanes
in the western Indian sector of the Southern Ocean and their relations with
geochemical tracers. Mar. Chem. 35, 151–167 (1991).
18. Meredith, M. P. et al. On the sources of Weddell Gyre Antarctic Bottom Water.
J. Geophys. Res. 105, 1093–1104 (2000).
19. Hoppema, M. et al. Prominent renewal of Weddell Sea Deep Water from a
remote source. J. Mar. Res. 59, 257–279 (2001).
20. Orsi, A. H. & Whitworth, T. III in Hydrographic Atlas of the World
Ocean Circulation Experiment (WOCE). Volume 1: Southern Ocean (eds
Sparrow, M., Chapman, P. & Gould, J.) (International WOCE Project Office,
2005).
21. Meijers, A. J. S., Klocker, A., Bindoff, N. L., Williams, G. D. & Marsland, S. J.
The circulation and water masses of the Antarctic shelf and continental slope
between 30 and 80◦ E. Deep-Sea Res. II 57, 723–737 (2010).
22. Middleton, J. H. & Humphries, S. E. Thermohalinestructure and mixing in the
region of Prydz Bay, Antarctica. Deep-Sea Res. 36, 1255–1266 (1989).
23. Nunes Vaz, R. A. & Lennon, G. W. Physical oceanography of the Prydz Bay
region of Antarctic waters. Deep-Sea Res. I 43, 603–641 (1996).
24. Yabuki, T. et al. Possible source of the Antarctic Bottom Water in the Prydz
Bay region. J. Oceanogr. 62, 649–655 (2006).
25. Tamura, T., Ohshima, K. I. & Nihashi, S. Mapping of sea-ice production for
Antarctic coastal polynyas. Geophys. Res. Lett. 35, L07606 (2008).
26. Fedak, M.A. Marine mammals as platforms for oceanographic sampling: A
win/win situation for biology and operational oceanography. Mem. Natl Inst.
Polar Res. 58, 133–147 (2004).
27. Charrassin, J-B. et al. Southern Ocean frontal structure and sea-ice
formation rates revealed by elephant seals. Proc. Natl Acad. Sci. USA 105,
11634–11639 (2008).
28. Williams, G. D. et al. Upper ocean stratification and sea ice growth rates during
the summer-fall transition, as revealed by Elephant seal foraging in the Adélie
Depression, East Antarctica. Ocean Sci. 7, 185–202 (2011).
29. Massom, R. A., Harris, P. T., Michael, K. J. & Potter, M. J. The distribution and
formative processes of latent-heat polynyas in East Antarctica. Ann. Glaciol. 27,
420–426 (1998).
30. Fraser, A. D., Massom, R. A., Michael, K. J., Galton-Fenzi, B. K. & Lieser, J. L.
East Antarctic landfast sea-ice distribution and variability, 2000–2008. J. Clim.
25, 1137–1156 (2012).
31. Matsumura, Y. & Hasumi, H. Modeling ice shelf water overflow and
bottom water formation in the southern Weddell Sea. J. Geophys. Res. 115,
C10033 (2010).
32. Darelius, E., Smedsrud, L. H., Østerhus, S., Foldvik, A. & Gammelsrød, T.
Structure and variability of the Filchner overflow plume. Tellus Ser. A 61,
446–464 (2009).
33. Wang, Q., Danilov, S. & Schröter, J. Bottom water formation in the southern
Weddell Sea and the influence of submarine ridges: Idealized numerical
simulations. Ocean Model. 28, 50–59 (2008).
34. Budillon, G., Castagno, P., Aliani, S., Spezie, G. & Padman, L. Thermohaline
variability and Antarctic bottom water formation at the Ross Sea shelf break.
Deep-Sea Res. I 58, 1002–1018 (2011).
35. Roquet, F. et al. Delayed-mode calibration of hydrographic data obtained
from animal-borne satellite relay data loggers. J. Atmos. Ocean. Technol. 41,
787–801 (2011).
36. Gordon, A. L., Visbeck, M. & Huber, B. Export of Weddell Sea Deep and
Bottom Water. J. Geophys. Res. 106, 9005–9017 (2001).
37. Orsi, A. H. & Wiederwohl, C. L. A recount of Ross Sea waters. Deep-Sea Res. II
56, 778–795 (2009).
38. Orsi, A. H., Smethie, W. M. Jr & Bullister, J. L. On the total input of Antarctic
waters to the deep ocean: A preliminary estimate from chlorofluorocarbon
measurements. J. Geophys. Res. 107, 3122 (2002).
39. Meredith, M. P., Watson, A. J., van Scoy, K. A. & Haine, T. W. N.
Chlorofluorocarbon-derived formation rates of the deep and bottom
waters of the Weddell Sea. J. Geophys. Res. 106, 2899–2919 (2001).
40. Naveira Garabato, A. C., McDonagh, E. L., Stevens, D. P., Heywood, K. J. &
Sanders, R. J. On the export of Antarctic Bottom Water from the Weddell Sea.
Deep-Sea Res. II 49, 4715–4742 (2002).
41. Kusahara, K., Hasumi, H. & Williams, G. D. Impact of Mertz Glacier Tongue
calving on dense shelf water. Nature Commun. 2, 159 (2011).
42. Tamura, T., Williams, G. D., Fraser, A. D. & Ohshima, K. I. Potential regime
shift in decreased sea ice production after the Mertz Glacier calving. Nature
Commun. 3, 826 (2012).
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.
239
NATURE GEOSCIENCE DOI: 10.1038/NGEO1738
ARTICLES
43. Meredith, M. P., Naveira Garabato, A. C., Gordon, A. L. & Johnson, G. C.
Evolution of the deep and bottom waters of the Scotia Sea, Southern Ocean,
during 1995–2005. J. Clim. 21, 3327–3343 (2008).
44. Fahrbach, E. et al. Warming of deep and abyssal water masses along the
Greenwich meridian on decadal time scales: The Weddell gyre as a heat buffer.
Deep-Sea Res. II 58, 2508–2523 (2011).
45. Purkey, S. G. & Johnson, G. C. Global contraction of Antarctic Bottom Water
between the 1980s and 2000s. J. Clim. 25, 5830–5844 (2012).
46. Harris, P. T. Ripple cross-laminated sediments on the East Antarctic Shelf:
Evidence for episodic bottom water production during the Holocene?
Mar. Geol. 170, 317–330 (2000).
the French Spatial Agency (CNES) and the French Polar Institute (IPEV). This work was
supported by Grants-in-Aids for Scientific Research (20221001, 20540419, 21740337,
23340135) of the Ministry of Education, Culture, Sports, Science and Technology in
Japan, and the Australian Government’s Cooperative Research Centres Programme
through the Antarctic Climate & Ecosystem Cooperative Research Centre.
Acknowledgements
Additional information
We are deeply indebted to the officers, crew and scientists on board TR/V Umitaka-maru
and R/V Hakuho-maru for their help with field observations. Comments by R. A. Massom
and support from K. Shimada and K. Kitagawa were helpful. The AMSR-E and SSM/I
data were provided by the National Snow and Ice Data Center (NSIDC), University of
Colorado. The ASAR data were provided by European Space Agency. IMOS seal CTD data
were provided through the Australian Animal Tracking and Monitoring System, a facility
of Integrated Marine Observing System. Isles Kerguelen deployments were supported by
240
Author contributions
K.I.O. and Y.F. conducted and analysed mooring observations after planning the
experiment with M.W., S.A. and T.T. G.D.W. led the investigation of seal data with
calibration by F.R., analysis by L.H-B., and fieldwork by I.F. and M.H. S.N. and T.T.
analysed satellite data. Y.K. and D.H. conducted hydrographic observations.
Supplementary information is available in the online version of the paper. Reprints and
permissions information is available online at www.nature.com/reprints. Correspondence
and requests for materials should be addressed to K.I.O.
Competing financial interests
The authors declare no competing financial interests.
NATURE GEOSCIENCE | VOL 6 | MARCH 2013 | www.nature.com/naturegeoscience
© 2013 Macmillan Publishers Limited. All rights reserved.